Anthropic’s Breakthrough Makes AI Understandable
Plus: Google Cloud's new AI capabilities for healthcare, SAP's AI innovations for spend management.
Hello Engineering Leaders and AI Enthusiasts!
Welcome to the 122nd edition of The AI Edge newsletter. This edition brings you Anthropic’s latest research that makes AI more understandable.
And a huge shoutout to our amazing readers! Your support is invaluable to us😊
In today’s edition:
🧠 Anthropic’s latest research makes AI understandable
🔥Google Cloud launches new generative AI capabilities for healthcare
🚀 SAP’s new generative AI innovations for spend management
📚 Knowledge Nugget: How Speculative Sampling Can Increase Your LLM's Inference Speed! by
Let’s go!
Anthropic’s latest research makes AI understandable
Unlike understanding neurons in a human’s brain, understanding artificial neural networks can be much easier. We can simultaneously record the activation of individual neurons, intervene by silencing or stimulating them, and test the network's response to any possible input. But…
In neural networks, individual neurons do not have consistent relationships to network behavior. They fire on many different, unrelated contexts.
In its latest paper, Anthropic finds that there are better units of analysis than individual neurons, and has built machinery that lets us find these units in small transformer models. These units, called features, correspond to patterns (linear combinations) of neuron activations. This provides a path to breaking down complex neural networks into parts we can understand and builds on previous efforts to interpret high-dimensional systems in neuroscience, ML, and statistics.
Why does this matter?
This helps us understand what’s happening when AI is “thinking”. As Anthropic noted, this will eventually enable us to monitor and steer model behavior from the inside in predictable ways, allowing us greater control. Thus, it will improve the safety and reliability essential for enterprise and societal adoption of AI models.
Google Cloud launches new generative AI capabilities for healthcare
Google Cloud introduced new Vertex AI Search features for healthcare and life science companies. It will allow users to find accurate clinical information much more efficiently and to search a broad spectrum of data from clinical sources, such as FHIR data, clinical notes, and medical data in electronic health records (EHRs). Life-science organizations can use these features to enhance scientific communications and streamline processes.
Why does this matter?
Given how siloed medical data is currently, this is a significant boon to healthcare organizations. With this, Google is also enabling them to leverage the power of AI to improve healthcare facility management, patient care delivery, and more.
SAP’s new generative AI innovations for spend management
SAP announced new business AI and user experience innovations in its comprehensive spend management and business network solutions to help customers control costs, mitigate risk, and increase productivity.
SAP will also embed Joule, its new generative AI copilot, throughout its cloud solutions, with availability in its spend management software planned for 2024. It has also unveiled SAP Spend Control Tower, which offers advanced AI features and the ability to see across all SAP spend solutions.
All these new AI innovations are being developed with security, privacy, compliance, ethics, and accuracy in mind.
Why does this matter?
This signifies SAP's commitment to revolutionizing every aspect of business through the power of generative AI. SAP is thoughtfully integrating cutting-edge AI into its market-leading solutions, ultimately helping customers achieve new levels of productivity and success.
Enjoying the daily updates?
Refer your pals to subscribe to our daily newsletter and get exclusive access to 400+ game-changing AI tools.
When you use the referral link above or the “Share” button on any post, you'll get the credit for any new subscribers. All you need to do is send the link via text or email or share it on social media with friends.
Knowledge Nugget: How Speculative Sampling Can Increase Your LLM's Inference Speed!
In the world of AI, bigger models and more data means better results. But as models get larger, questions arise about how to handle them effectively.
In this interesting article,
focuses on how can we improve LLM’s inference speed. To answer, he digs deeper to see if it’s even an issue in the first place, if it is what is causing it to be slow and how can we solve it.
Using a small draft model we can sample tokens that allow us to create a bigger batch and pass them to a bigger target model to approve or reject. This way we are able to skip many forward passes on easy to predict tokens.
Why does this matter?
As NLP models grow larger and we introduce the word Large in language models, we are stepping into more engineering challenging fields– how do we scale such big models? how do we train with this many GPUs? How do we optimize training speed? and so on. This article explores improving LLMs’ inference speed which is crucial in optimizing them for practical use.
What Else Is Happening❗
📱ChatGPT’s mobile app hit record $4.58M in revenue last month, but growth is slowing
This gross revenue is across its iOS and Android apps worldwide. But while it was topping 31% in July and 39% in August, that dropped to 20% growth as of September. (Link)
🚀Mendel launches AI-Copilot for real world data applications in healthcare
Called Hypercube, it enables life sciences and healthcare enterprises to interrogate their troves of patient data in natural language through a chat-like experience. It can deliver blazing-fast insights and answer previously unanswerable questions. (Link)
💰Lambda Labs, AWS’s competitor offering servers with Nvidia chips, nears $300M funding
Like AWS, it rents out servers with Nvidia chips to AI developers and is nearing a $300M equity financing. As demand for Nvidia’s AI chips has skyrocketed this year, revenue at startups such as Lambda has boomed, attracting investors. (Link)
🌾AI drones successfully monitor crops to report the ideal time to harvest
For the first time, researchers have demonstrated a largely automated system that heavily uses drones and AI to improve crop yields. It carefully and accurately analyzes individual crops to assess their likely growth characteristics. (Link)
🌍Scientists achieve 70% accuracy in AI-driven earthquake predictions
An AI tool predicted earthquakes with 70% accuracy a week in advance, as observed during a 7-month trial held in China. Based on its analysis, the tool successfully anticipated 14 earthquakes. This promising experiment was conducted by researchers from The University of Texas (UT) at Austin, USA. (Link)
That's all for now!
Subscribe to The AI Edge and join the impressive list of readers that includes professionals from Moody’s, Vonage, Voya, WEHI, Cox, INSEAD, and other reputable organizations.
Thanks for reading, and see you tomorrow. 😊